Subduction related tectonic evolution of the ...

Precambrian Research 227 (2013) 204–226

Contents lists available at SciVerse ScienceDirect

Precambrian Research journal homepage: www.elsevier.com/locate/precamres

Subduction related tectonic evolution of the Neoarchean eastern Dharwar Craton, southern India: New geochemical and isotopic constraints M. Ram Mohan a,∗ , Stephen J. Piercey b , Balz S. Kamber c , D. Srinivasa Sarma a a b c

CSIR – National Geophysical Research Institute, Hyderabad 7, India Department of Earth Sciences, Memorial University of Newfoundland, St. Johns, NL A1B 3X5, Canada School of Natural Sciences, Trinity College Dublin, Dublin 2, Ireland

a r t i c l e

i n f o

Article history: Received 1 November 2011 Received in revised form 9 June 2012 Accepted 20 June 2012 Available online 3 July 2012 Keywords: Dharwar Craton Geochemistry Radiogenic isotopes Crustal recycling Subduction

a b s t r a c t The Neoarchean eastern Dharwar Craton (EDC) is distinct from the Mesoarchean western Dharwar Craton (WDC) in many aspects of its geology. The important distinguishing features of the EDC are the predominance of younger granitoids, abundance of gold mineralization and the exposure at lower crustal depths. No consensus exists on evolutionary models for EDC; mutually exclusive plume and subduction-derived tectonic models have been proposed. Geochemical and radiogenic isotopic studies on the granitoids and volcanic rocks of three greenstone belts along a cross-section in the northern part of EDC are presented herein. The evolved Nd isotopic signatures, radiogenic Pb isotopic ratios and “arc-like” geochemical signatures are suggestive of a subduction regime and the involvement of recycled older crust in the derivation of these rocks. The proposed petrogenetic mechanism involves multi stage processes in a supra subduction regime involving slab dehydration, formation of hydrous basaltic melts, and re-melting and interaction with sub-arc basaltic crust at low pressures where amphibole ± plagioclase is the dominant residual phase. There is a notable systematic decrease in the extent of older crustal involvement from west to east in the EDC. This is in concurrence with the younging of Dharwar Craton from west to east and eastward subduction. The proposed petrogenetic model can efficiently explain the variations in older crustal involvement, which is very common in other Archean cratons. © 2012 Elsevier B.V. All rights reserved.

1. Introduction The Neoarchean represents a key period of continental crustal growth with major tectonic assembly and stabilization of large cratonic areas such as the Superior, Yilgarn, and Dharwar cratons (Condie, 2000; Condie and Kröner, 2012). These cratons are composed predominantly of granite-greenstone successions and tonalite–trondhjemite–granodiorite (TTG) terranes, evolved through various geodynamic processes that were likely the product of combinations of subduction zone and plume tectonics (Abbott, 1996; Condie, 1994; Huang et al., 2012; Naqvi et al., 2006; Polat et al., 2011; Polat and Kerrich, 2001; Zhai and Santosh, 2011). Though there is considerable understanding of the crustal growth and assemblages of most of these cratons, studies integrating the temporal and petrogenetic evolution of the volcanic and plutonic rocks are very limited for many cratons, particularly so for the Dharwar Craton. The Dharwar Craton represents one of the largest cratonic masses of the Indian Shield. It is divided into two principal terrains:

∗ Corresponding author. Tel.: +91 4023434700; fax: +91 4023434651. E-mail address: [email protected] (M.R. Mohan). 0301-9268/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.precamres.2012.06.012

the Eastern Dharwar Craton (EDC) and the Western Dharwar Craton (WDC) (Fig. 1). The EDC is different from WDC in many ways. The WDC contains mostly older gneissic rocks (3.3–2.7 Ga) and thicker crust, the EDC has thinner crust (Gupta et al., 2003a) and is comprised of predominantly younger juvenile granitoids (∼2.52 Ga) (Chardon et al., 2011, 2002; Dey et al., 2012; Jayananda et al., 2000), and auriferous greenstone belts (2.5–2.7 Ga) that are surrounded by variably evolved syn-kinematic granitoids (Balakrishnan and Rajamani, 1987; Naqvi, 2005; Rajamani et al., 1987). Available geochronological data for the rocks of EDC suggest that greenstone volcanism, granitoid formation, accretion and deformation occurred between 2.7 and 2.5 Ga (Chardon and Jayananda, 2008; Chardon et al., 2011; Jayananda et al., 2012; Nutman et al., 1996; Rogers et al., 2007; Sarma et al., 2012). Like other Archean cratons the evolution and assembly of the EDC has been the subject of debate with the main hypothesis favoring a plume growth model (Jayananda et al., 2000), whereas other authors favor a subduction model (Chadwick et al., 2000), or a combined model involving the combination of lateral and vertical tectonics (Manikyamba and Kerrich, 2012; Naqvi et al., 2002, 2006). The southern part of the EDC is relatively well-studied because of the presence of two well-known gold deposits (Kolar and Ramagiri). Based on the isotopic and geochemical characteristics on these

M.R. Mohan et al. / Precambrian Research 227 (2013) 204–226

205

Fig. 1. Generalised geological map of southern Indian Peninsula. Study areas are shown in boxes.

rocks, a convergent margin setting has been proposed for its evolution (Balakrishnan et al., 1990; Zachariah et al., 1996). Geochemical and isotopic studies on the younger granitoids of southern EDC through a west to east cross-section from the Closepet granite to the eastern margin of the Kolar greenstone belt, have led workers to propose generation of these rocks from a heterogeneous mantle with predominantly juvenile additions (Jayananda et al., 2000). A depleted mantle source has been proposed for the mafic volcanic rocks of EDC (Manikyamba et al., 2004a). Some of the felsic volcanic rocks and granitoids from the northern greenstone belts of EDC have geochemical systematics similar to that of modern “adakites” and slab melting processes are interpreted to have been important in their genesis (Manikyamba et al., 2008, 2009; Naqvi et al., 2006, 2008; Prabhakar et al., 2009). In this paper, geochemical and isotopic data are provided for syn-kinematic granitoids, and stratigraphically well constrained felsic and mafic volcanic rocks from three major Neoarchean greenstone belts of the northern EDC, namely the Sandur, Kushtagi and Hutti belts (Fig. 1). This study also reports new major and trace element data for all rock types, as well as common Pb and Nd isotopic data for some of the felsic magmatic suites. The latter data will be used with trace element data to quantify potential crust–mantle interactions, to test whether these rocks contain juvenile signatures and record juvenile crustal additions similar to the younger granitoids from the southern part of EDC (Jayananda et al., 2000), and whether or not the mantle composition is uniform throughout

the EDC. The study also aims to understand the tectonic processes responsible for the petrogenesis and evolution of the EDC. Furthermore, adakite-like felsic volcanic and granitoid rocks are present in all these three greenstone belts (Naqvi et al., 2008), and an attempt is made herein to test whether these granitoids and felsic “adakitic” rocks display true slab melt signatures similar to modern adakites (Stern and Kilian, 1996). 2. Regional geological and tectonic framework The Dharwar Craton (DC) exposes a large section of Archean continental crust (Fig. 1). Formation and the growth of different rock suites of DC appear to have taken place in more than one billion years between 3.6 and 2.5 Ga (Radhakrishna and Naqvi, 1986). The Dharwar Craton is divided into two distinct tectonic blocks: the Western Dharwar Craton (WDC) and Eastern Dharwar Craton (EDC) (Naqvi and Rogers, 1987; Swaminath and Ramakrishnan, 1981). Swaminath et al. (1976) marked an arbitrary boundary between both the blocks parallel to the western margin of Closepet granite, which was subsequently modified to coincide with the Chitradurga boundary fault and confirmed by deep seismic sounding (Kaila et al., 1979) and by Landsat imagery (Drury and Holt, 1980). Based on the recent structural, deformational and geochronological studies, it is widely agreed that thrust zone found along the eastern margin of Chitradurga greenstone belt is the boundary between both the blocks (Chadwick et al., 2003; Chardon and

206


Jayananda, 2008; Chardon et al., 2011; Jayananda et al., 2006). The WDC is largely made of older Peninsular Gneissic Complex (PGC) (3.3–2.7 Ga) (Beckinsale et al., 1980; Bhaskar Rao et al., 1992; Meen et al., 1992) and large greenstone belts that are predominantly composed of metasedimentary rocks and komatiite–tholeiitic volcanic suites (Jayananda et al., 2008; Peucat et al., 1995), whereas the EDC is dominated by calc-alkaline granitoids that are interspersed with thin and linear greenstone belts (Jayananda et al., 2000; Naqvi, 2005). Based on the close lithological similarity, structural coherence and emplacement between 2.7 and 2.5 Ga, the plutonic complex of EDC was termed as Dharwar Batholith (Chadwick et al., 2000). A summary of the comparison between both the blocks can be found in Table 1 of Naqvi and Prathap (2007). The study area is located in the northern part of EDC that constitutes a horizontal cross-section, passing from Sandur greenstone belt (SGB; Fig. 2A) in the west, through Kushtagi greenstone belt (KGB; Fig. 2B), into the Hutti-Maski greenstone belt (HMGB; Fig. 2C) in the east (Fig. 1).

2.1. Sandur greenstone belt The volcano-sedimentary sequences of Sandur greenstone belt (SGB) (Fig. 1) overlaps the western marginal zone of Dharwar batholith surrounded by the mixture of older PGC (>2900 Ma) and younger anatectic granites (∼2500 Ma) (Chadwick et al., 1996). Early work on this belt suggested that it was a predominantly conformable stratigraphic sequence (Chadwick et al., 2000; Roy and Biswas, 1983), whereas other workers suggest that its tectonostratigraphy is a result of terrane accretion (Manikyamba and Kerrich, 2006; Naqvi et al., 2002). The SGB contains a diverse assemblage of tectonically juxtaposed volcanic rocks, which includes submarine pillowed low K-tholeiitic to high Mg-basalts, ultramafic schists and calc-alkaline rhyolites along with an assemblage of terrigenous and chemical sediments that includes metagreywacke, chert, banded iron formation (BIF), quartzite and shale. The high temperature-low pressure metamorphism (HT/LP) of this belt was inferred to be contemporaneous with regional deformation and granite emplacement (Chadwick et al., 1996; Roy and Biswas, 1983). The volcanic rocks for this study were collected from the Vibhutigudda Formation (Prasad et al., 1997), which is also known as eastern felsic volcanic terrain (EFVT) of the SGB (Fig. 2A; Manikyamba et al., 2008). The EFVT consists of metabasalt, felsic volcanic rocks, greywacke-carbonaceous shale – polymictic conglomerate and BIF (Figs. 2A and 3; Subba Rao et al., 2001). Greywacke of the EFVT have island arc provenance and the associated carbonaceous shales are interpreted as the first cycle volcanogenic sedimentary rocks deposited in an intraoceanic arctrench complex (Manikyamba and Kerrich, 2006). Zircons from the felsic volcanic rocks of the EFVT have yielded Concordia upper intercept age of 2691 ± 18 Ma and 2658 ± 14 Ma for two samples (Nutman et al., 1996). The available SHRIMP U–Pb zircon ages for the granitoids surrounding Sandur belt (2719 ± 40 Ma and 2570 ± 62 Ma) suggests prolonged Neoarchean emplacement history (Nutman et al., 1996). Based on the structural relationships, the younger Rb-Sr whole rock age of Toranagallu granite (2452 ± 50 Ma) (Bhaskar Rao et al., 1992) is interpreted as the cooling age (Nutman et al., 1996). These granitoids range from banded migmatitic gneisses to porphyritic granites, have intrusive contacts parallel to the schistosity of the belt. Based on the relative ages and deformation, these granitoids are divided into six types (Chadwick et al., 1996). Granitoids for this study were collected from working quarries from the Papinayakanahalli granite in the north and Toranagallu granite in the southern part of the belt. These are homogeneous grey granites, coarse-medium grained with enclaves

of porphyritic pink-grey granite at places, however the sampling is avoided from such enclaves. 2.2. Kushtagi greenstone belt The Kushtagi greenstone belt (KGB) occurs in the central part of the EDC and is the northern continuation of the 400 km long Ramagiri-Hungund composite greenstone belt (Fig. 1). A shear zone passes through the centre of the belt along which numerous gold occurrences are found (Manikyamba et al., 2004b). Structural studies reveal two deformational events for this belt and the emplacement of surrounding granitoids was related to late phase of D1 transpression (Matin, 2006). No age information is available for the rocks of this belt, however U–Pb zircon and titanite ages are available for the felsic volcanics and surrounding granitoids of Ramagiri belt (Balakrishnan et al., 1999), which is the southern end of Ramagiri-Hungund composite greenstone belt (Fig. 1). The emplacement age for felsic volcanics of the Ramagiri belt is 2707 ± 18 Ma and the granites surrounding the belt have the emplacement ages in the range between 2650 ± 7 Ma and 2595 ± 1 Ma (Balakrishnan et al., 1999), same crystallization ages are considered for the Kushtagi rocks for Nd and Pb isotopic calculations (Table 2). The major lithologies of this belt are metabasalt, metarhyolite and minor BIF (Figs. 2B and 3). The metabasalts are generally massive and pillowed, consist of plagioclase and pyroxene, altered to chlorite–actinolite at places. The rhyolites contain quartz, soda-rich plagioclase and traces of K-feldspar-sericite. Geochemical signatures of the metabasalts suggest high-Mg tholeiitic nature, whereas the felsic volcanic rocks are comparable to adakites (Naqvi et al., 2006) The interlayered basic and felsic volcanic rocks were collected from the central part of the Kushtagi belt, while the granitoids are collected from either side of the belt (Fig. 2B). The granitoids surrounding the belt are generally grey granite, display sub-vertical NW trending foliation and their intrusive relationship with the belt suggest syn-kinematic emplacement. Older gneissic enclaves are seen at places within the granitoids and sampling is avoided from such enclaves. Recent geochemical and Nd isotopic studies on the granitoids surrounding this belt suggest distinct crustal histories on either side of the belt and subsequent juxtaposition due to lateral accretion (Dey et al., 2012). 2.3. Hutti-Maski greenstone belt The Hutti-Maski greenstone belt (HMGB), located in the northern part of the EDC, is currently the sole primary gold producing greenstone belt in India with an annual gold production of ∼3.0 tons (Mohan and Sarma, 2010). Unlike the other linear greenstone belts of EDC, the HMGB is a horseshoe-shaped, elongate belt (Fig. 1). The major lithological units are pillowed and massive metabasalt, intermediate to felsic volcanic rocks and quartz-phyric rhyolite porphyry (Figs. 2C and 3). Roy (1979) recognized three generations of folds (F1, F2 and F3), with the intrusion of granodiorite between F1 and F2. The superposition of these three fold generations is interpreted to have produced the upright cuspate structure of the greenstone belt (Roy, 1991). The western boundary of the belt is tectonically juxtaposed against the older TTG basement, whereas the northern, southern, and eastern margins are intruded by younger granitoids. Gold mineralization is controlled by steeply dipping, NNW–SSE trending strike slip shear zones locally referred to as reefs (Naganna, 1987). Based on the alteration mineralogy and structural studies, the gold mineralisation is related to two tectonic events (Kolb et al., 2005). The felsic volcanic rocks have yielded a U–Pb zircon crystallization age of 2587 ± 7 Ma, whereas the adjoining Kavital granitoid has yielded a U–Pb zircon age of 2545 ± 7 Ma (Sarma et al., 2008). A titanite U–Pb age of 2532 ± 3 Ma has also been reported for the Yelagatti granitoid (Anand and Balakrishnan,


207

Fig. 2. (A) Simplified geological map of Eastern Felsic Volcanic Terrain (Subba Rao et al., 2001). Inset shows the disposition of EFVT in the outline map of Sandur greenstone belt. (B) Simplified geological map of the central part of Kushtagi greenstone belt (Naqvi et al., 2006). (C) Geological map of Hutti greenstone belt (Srikantia, 1995).

2010). Recent U–Pb zircon age of 2569 ± 13 Ma for the felsic volcanic rock from Buddine area further confirms the predominance of 2.57 Ga event in this area (Jayananda et al., 2012). The least altered basic and felsic volcanic rocks were collected from the central part of the belt, while two altered samples are from the 24th level of Hutti gold mine (Fig. 2C). The granitoids are from the Kavital and Yelagatti (Fig. 2C). 3. Petrographic characteristics Most of the granitoids were collected from working quarries, whereas fresh volcanic rocks lacking visible veins and alteration were sampled from outcrops. Two alteration zone samples from the Hutti underground gold mine were studied to understand the effect of alteration, but the data of these two samples are not used for any petrogenetic interpretations. The sample suite provides a

comprehensive stratigraphic and temporal distribution of major lithologic units in the northern part of EDC. Sandur granitoids are chiefly composed of quartz, K-feldspar (microcline and orthoclase), and plagioclase feldspar (Fig. 4A). Biotite is the dominant mafic phase. Chlorite, sphene, epidote and zoisite are accessory minerals. Quartz and feldspar are medium grained and mostly anhedral. The granitoids of Kushtagi and Hutti have more relative abundance of plagioclase feldspar over potash feldspar (orthoclase). Amphibole, biotite, sphene and opaques are the other constituting minerals. Quartz is anhedral whereas plagioclase and K-feldspars are subhedral. Plagioclase exhibits zoning (Fig. 4B), often display polysynthetic twinning. Deformation is common and can be inferred from the breaking of feldspar grain and recrystallization and grain boundary migration of quartz. Plagioclase is saussuritized (Fig. 4C), where the cores are more saussuritized than rims in zoned plagioclase crystals. Biotite, chlorite

208


Fig. 3. Stratigraphic columns for the EFVT of Sandur, Kushtagi and Hutti greenstone belts having older units at the bottom. Age data as described in the text.

and epidote are other constituting minerals; titanite and magnetite are common accessories (Fig. 4D). The Sandur rhyolites are made up of quartzo-feldspathic matrix; the primary minerals being quartz, plagioclase and muscovite, often exhibits porphyritic texture. The rhyolites from the central part of Kushtagi belt are sheared resulting in the granulation of quartzo-feldspathic matrix and rotation of plagioclase phenocryst at various angles to the shear plane (Fig. 4E). Plagioclase is euhedral to subhedral showing well developed twinning and zoning. Quartz also occurs as phenocryst as well in the form of veins, while the plagioclase phenocrysts display polysynthetic twinning (Fig. 4E). Quartz is generally anhedral, display cracks that are filled with secondary minerals during deformation. The Hutti rhyolites show S–C deformational fabrics and the development of quartz eye (Fig. 4F). The basic rocks are generally medium- to coarse-grained amphibolites, primarily made of hornblende, intergrown with chlorite, plagioclase and quartz. 4. Geochemistry

Supplementary Table 1. AGV-2 and BHVO-2 were analyzed as unknowns, the long-term reproducibility of which is reported in Supplementary Table 1. For the isotope analysis, the remaining solution of the rock digest after trace element analysis, constituting more than 98% of the original solution, was dried down and converted with HBr. Lead was purified on miniaturized anion exchange columns according to the method outlined in Kamber and Gladu (2009). Lead isotope ratios were determined by ICP-MS according to the protocol of Ulrich et al. (2010). The column eluate from the anion column was then passed over a cation column on which the REE fraction was obtained. This was passed over an LN spec column to yield a Sm-free Nd separate, which was analyzed for isotope composition using a Finnegan Triton thermal ionization mass spectrometer (TIMS) at Carleton University. No isotope dilution analysis of the Sm/Nd ratio was necessary due to the high precision and accuracy of the employed trace element protocol (Babechuk and Kamber, 2011).

4.1. Analytical techniques

4.2. Alteration

All the samples were initially reduced in a jaw crusher, and then manually powdered in agate mortars. Major elements were analyzed by X-ray fluorescence spectroscopy (XRF) at the Ontario Geoscience Laboratories (GeoLabs), Sudbury, Ontario, Canada using fused discs. Trace elements, including the high field strength elements (HFSE) and rare earth elements (REE) were analyzed using inductively coupled plasma mass spectrometry (ICP-MS) Thermo X Series II ICP-MS instrument in the Department of Earth Sciences, Laurentian University. Sample preparation for trace element analysis involved the bomb digestion of 100 mg of sample for 72 h in HF-HNO3 at 195 ◦ C. Fluorides were converted with HCl, followed by HNO3 . Solutions were run at 1:3000 with matrix tolerant cones at low extraction voltages. 6 Li, In, 147 Sm, Re, Bi and 235 U were used as internal standards. External drift was corrected by repeated analyses of a 1:2500 solution of BHVO-2. Instrument response was calibrated relative to two independent digestions of W-2, the preferred concentrations of which are given in the

Various geochemical studies have established that ‘immobile’ elements such as Al2 O3 , TiO2 , Zr, Y, Nb, Ta, Sc and REE (except “Ce” and “Eu”) are least sensitive to hydrothermal alteration and metamorphism commonly found in Archean terrains, and therefore give insight into the primary petrogenetic processes that formed rocks in ancient terrains (Humphris, 1984; Jochum et al., 1991; Piercey et al., 2002; Polat and Kerrich, 2000). In contrast, mobile elements and their ratios provide insights into the effects of alteration in Archean rocks (Kamber et al., 2002; Mohan et al., 2008). It has been shown in many studies that Zr (and other HFSE and REE) is one of the least mobile elements during regional metamorphism and hydrothermal alteration, and is mobile only under very specific conditions (e.g., acid-sulfate conditions; presence of F) (Pearce and Peate, 1995; Winchester and Floyd, 1977). Zirconium is therefore a very useful index upon which other elements can be compared to test for immobility (e.g., Fig. 5). Maclean (1990) and MacLean and Barrett (1993) had illustrated that when elements are plotted

Table 1 Major and trace elemental compositions for various rocks of EDC. RB-4 Sandur Rhyolite

RB-6 Sandur Rhyolite

RB-15 Sandur Rhyolite

RBK-25 Kushtagi Rhyolite





RBH-45 Hutti Rhyolite




RB-9 Sandur Granitoid



KN-120 Sandur Granitoid

Northing Easting SiO2 (wt.%) TiO2 Al2 O3 Fe2 O3 MgO MnO CaO K2 O Na2 O P2 O5 LOI Total Mg#

15.08.21.0 76.46.05.2 77.25 0.28 13.12 0.34 0.24 0.01 2.86 2.50 3.16 0.10 0.98 100.84 60.84

15.08.22.9 76.46.07.1 77.45 0.31 12.89 0.27 0.14 0.02 2.00 1.34 4.97 0.10 1.11 100.60 53.30

15.06.14.6 76.47.47.6 71.50 0.31 14.28 2.38 0.77 0.04 2.05 3.65 3.90 0.10 1.26 100.24 41.59

15.11.11.2 76.42.46.7 73.16 0.38 14.27 0.55 0.39 0.02 4.63 2.55 2.92 0.20 1.87 100.94 60.95

15.49.08.2 76.25.44.1 71.73 0.28 14.89 2.23 0.82 0.03 2.85 0.84 5.26 0.10 1.39 100.42 44.73

15.15.27.0 76.26.50.9 75.59 0.15 13.46 1.75 0.21 0.03 1.60 3.93 3.78